Spatial proximity of the two essential sulfhydryl ... - ACS Publications

Sep 10, 1974 - ine and Kielley, 1964; Yamaguchi and Sekine, 1966), partici- pate in the formation of a cyclic ternary complex withMgATP. (Burke et at...
0 downloads 0 Views 446KB Size
Biochemistry @ Copyright 1974 by the American Chemical Society

Volume 13, Number 19

September 10, 1974

Spatial Proximity of the Two Essential Sulfhydryl Groups of Myosin? Emil Reisler,’ Morris Burke,§ Sylvia Himmelfarb, and William F. Harrington*

ABSTRACT: A bifunctional analog of N E M , p-phenylenedimaleimide, shows specificity similar to the monofunctional reagent toward the two essential sulfhydryl groups of myosin, S H , and SH2. The dimaleimide reagent, which has a crosslinking span of 12- 14 A, can bridge between these thiol groups

\415

have recently proposed that the two essential sulfhydryl groups of myosin, generally referred to as S H I and SH2 (Sekine and Kielley, 1964; Yamaguchi and Sekine, 1966), participate i n the formation of a cyclic ternary complex with MgATP (Burke ef ai.,1973; Reisler et al.. 1974) and that the stability of this complex is responsible for the well-known inhibition of A T P hydrolhsis i n the presence of 1 mM Mg2+. Mandelkow and h4andzlkow (1973) have proposed that the myosinMgATP complex is in the form of enzyme-metal-substrate, rather than enzyme-substrate-metal based on recent fluorometric studies. Their studies also reveal independence of the two “heads” in the interaction with MgATP. Actin apparently activates the MgATPase of myosin by binding at, or in the vicinity of, the S H ] site, thus disrupting the cyclic complex at this site (Burke et ai.,1973; Reisler et al., 1974; see also Seidel, 1973). Evidence supporting this mechanism for activation of A T P hydrolysis was obtained from a study of the influence of various ligands on the reaction between N E M 1 and myosin (Reisler et al., 1974). In this communication we wish to demonstrate that the two essential sulfhydryl groups of myosin, SHi and SH2. participating in the proposed cyclic ternary complex with MgATP are Contribution N o . 786 from the Department of Biology, McCollum-Pratt Institute, T h e Johns Hopkins University. Baltimore, Maryland 2 I2 18. Received April 10, 1974. This investigation was supported by U. S. Public Health Service Grant No. A M 04349. M. B. gratefully acknowledges a Muscular Dystrophy Associations of America postdoctoral fellowship. E. R. thanks the U. S. S t a t e Department for support under the Johns Hopkins-Weizmann Institute, Rehovot. Israel, Exchange Program. Present address: The Weizmann Institute of Science, Rehovot, Israel. § Present address: Department of Phjsiology, University of Maryland School of Dentistry, Baltimore, Md. 2 l 2 0 l . I Abbreviation used is: N E M , N-ethylmaleimide.

*

thus establishing their steric proximity. This finding provides experimental support for a proposed mechanism of A T P hydrolysis which requires that these two S H groups be in appropriate spatial arrangement to form a cyclic ternary complex with the metal substrates.

in close proximity to each other during their interaction with the Mg-nucleotide complex. Steric proximity of these groups is a prerequisite for formation of the proposed cyclic complex with MgATP and the results to be presented below show that a bifunctional reagent, N,N’-p-phenylenedimaleimide,with a cross-linking span of 12-14 A and a specificity for S H l and SH2 groups of myosin similar to that of N E M , can bridge between these groups. Materials and Methods A T P of highest available purity was purchased from Sigma Chemical Co., St. Louis, Mo. Inorganic salts and reagents were analytical grade, N E M was the product of Eastman Kodak Co. (Rochester, N . Y . ) and p-phenylenedimaleimide was obtained from Aldrich Chemical Co. The preparation of myosin has been described elsewhere (Godfrey and Harrington, 1970). S H I - N E M and SH2-NEM modified myosins were prepared as described earlier (Reisler et ai., 1974). Reaction of myosin with N,N’-p-phenylenedimaleimide(dissolved in acetone) was carred out in exactly the same way as the modification with N E M . The effect of these modifications on ATPase acitivty of myosin was tested as reported by Sekine and Kielley ( 1 964). Purification of S H I-phenylenedimaleimide myosin to remove the unreacted bifunctional reagent was achieved with a Sephadex (3-25 column equilibrated with pH 7.0 solvent (0.5 M KCI-0.05 M Tris). Myosin samples reacted for 30 min at pH 7.0 with an eightfold molar excess of the bifunctional reagent over protein were loaded on the column and eluted within 15 min. The loss in CaATPase activity was followed as a function of time at pH 7.0, 7.9, and in the presence of 1 mM MgADP (pH 7.9). All modifications were carried out at 5’. The Ca and EDTA activities of myosin were measured a t 37’ employing the procedures of Kielley and Bradley (1956) and Kielley er ai.( I 956). BIOCHEMISTRY, VOL.

13,

N O

19, 1974

3837

1

I

!

'>A\

'e>...-------a --

ob

10

--

E D T A ATFcse

----a ~

1

J

---

..

20

30

0

F I G U R E I: Relative ATPase activities of injosin reacted i i t pH 7.0 w i t h a n eightfold excess of \ E M ( - - - - 1 or p-phenqlencdimnleimide

(-)

2@

40

60

Time ( r n i n )

Time (mlnj

plotted as a function of time of modification.

t I L L K F 3: Relative Ca?+ A T P a x activitq of SH.-NEM modified m y osin reacted with :in eightfold molar excess of p-phen) Icncdimnlcirnidc a t pH 7.0 ( O ) , pH 7.9 (A),and pH 7.9 i n thc presence of 1 nihl M g A T P ( 0 ) .IOOYo ATPase activity of SH:-AEM mqosin correspond\ to 0.95 Mmol of P, mg-l m i n - I .

Results and Discussion The reaction of myosin with N E M has been characterized in detail over the past years both with respect to its catalytic consequences and the nature of the modified site on the protein (Sekine er ai., 1962; Sekine and Kielley, 1964; Yamashita rr nl., 1964; Yamashita et al., 1965: Yamaguchi and Sekine, 1966; Reisler et al., 1974). A clear correlation has been established between the titration of the two essential sulfhydryl groups S H 1 and SH2 with N E M and the observed changes in the activity properties of myosin. Blocking of the S H I group of myosin with N E M results in a marked increase i n CaATPasc and a decrease in EDTA ( K + ) ATPase activities (Sekine et ol.. 1962). Subsequent modification of the second essential sulfhjdry1 group, SH2, leads to a complete loss of both activities (Sekine and Yamaguchi, 1963; Yamaguchi and Sekine, 1966). W e have employed these changes in ATPase activities of mqosin in the present study to establish that a bifunctional analog of N EM, N,N-p-phenylenedimaleimide, reacts with myosin i n a similar fashion to the monofunctional h E M . Figure 1 shows that under conditions employed for modification of S H I by N E M (pH 7.0. 8.1 molar ratio of N E M to myosin) the reaction with p-phenylenedimaleimide results in a concomitant increase in CaATPase and decrease in EDTA ( K + ) ATPase activities. The rate and extent of these changes are som ew ha t s ma I1er with p - p h e n y I en ed i ma I e i m ide than tv it h N E M , possibly due to a lower reactivitj of the former toward S H 1 groups of myosin. The overall effect. however. is the same. Evidence that the bifunctional maleimide can react with both S H I and SHz groups was obtained from modification with this reagent of prelabeled S H l - N E M and S H z - N E M niy-

jl: Ih \

b - A

0A

-a _ _ 20

A

A

~

40

60

Time (rnln)

Relative Ca'+ ATPase activity o f S H , - r \ l E M modified m j osin reacted with p-phenLlenedimaleiniide ( p P D M ) . 100% ATPase ;ICtivity of S H l - N E M myosin corresponds to 2 pmol of P, mg-' m i n - ' . pH 7.0, 4-(O) and 8 - ( 0 )fold molar CXC'I";~ pPDM/niqosin: pH 7.9. ( A ) fourfold molar excess pPDM/m)osin; pH 7.9 ( A ) I m ~ Vi p 4 D P . fourfold molar excess p P D M . F I G L R ~ 2: z

3838

B I O C H E M I S T R Y .VOI

13.

N O

19. 1974

osin samples (prepared as described bq Reisler P I al,, 1974). Figure 2 demonstrates that at p H 7.0 and a t four- and eightfold molar excess of the reagent over protein, p-phenqlcnedimalcimide, does not affect the active site of S H , - U E M myosin. When the p H is raised to 7.9 some loss in CaATPasc i b observed; this loss is complete and rapid when the modification is performed in the presence of I mM MgADP. Thus the reaction pattern follows very closely the reaction of N E M with S H , Y E M myosin, as originally observed by Yamaguchi and Sekine (1966). The effect of p-phenylenedimaleimide on S H 2 N E M myosin, shown in Figure 3, resembles again the reaction of N E M with the SHz-labeled protein (Seidel, 1969: Reisler P I ai., 1974). The modification. a s followed by a loss in CaATPase activity, proceeds quite effectively a t ptl 7.0, is accelerated a t pH 7.9 and even more so in the presence of MgADP. This pattern of modification. however, differs significantly from the blocking of SHI-NEM myosin (Figure 7 ) . The results presented in Figures 1-3 establish the similarity between the reaction of N E M and p-phenqleneditnaleiinide with myosin and indicate that both reagents exhibit comparable specificity toward S H I and SH2 groups of the protein. We have observed previously (Reisler rt al.. 1974) that modification of S H I groups of myosin bq ;\EM can be achieved inore readily a t pH 7.9 than at p H 7.0. When modification i b conducted (at pH 7.9) ivith p-phenqlenedinialeimide the rapid initial elevation of CaATPase is followed by a SIOM decrcaw in this activity (Figure 4 ) . This behavior is unlike that obscrvcd ~ i t N h E M (upper dashed line in Figure 4),where the clcvated CaATPase activity remains invariant nith time (-60 min). A decrease in CaATPase activity is normally observed upon niodification of the S H 2 group subsequent to a prior blocking of'the S H I group (Yamaguchi and Sekinc. 1966). The initial decrease i n EDTA ( K + ) ATPase activity. shown i n Figure 1,is ;I nieiisure of myosin molecules which have reacted with the bil'unctional reagent. Within a minutes EDTA ATPasc reaches ;I constant value depending on the molar cxccss of the rcagent ovcr protein. ,Apparcntl\ a t this stage the reagcnt has been exhausted in competitive reactions. The decrcaw i n CaATPase activity mentioned above is confined to those molecules ivhich have been initiallj modified a t S H , since thc EDTA ATPase remains unaltered ovcr the examined timc interval. Had the loss i n CaATPasc bcen due t o a sloher reaction of S H - groups on previously unmodified niqosin niolcculc'i an additional decrease in EDT,A ATPase would have bcen observed (Seidel. 1969; Reisler rf ul.. 1971). The behavior of Ca and EDTA ATPases in Figure 4 cannot. h u u e v e r . exclude a hypothetical tno-step process: blocking o f

SULFHYDRYL GROUPS OF MYOSIN

A A -

0

20

40

60

80

I

0

10

20 30 Tlme [min)

F I G U R E 5: Loss in Ca2+ ATPase activity of SHI-p-phenylenedimaleimide myosin purified on Sephadex G-25 column plotted as a function of time. 100% ATPase activity of SH I-p-phenylenedimaleimide corresponds to 1.8 pmol of P, mg-’ min-I. Experiments carried out a t pH 7.0 (0); pH 7.9 (0); and pH 7.9, 1 mM M g A D P (A).

Time (rnin) F I G U R E 4: Relative C a 2 + ATPase (0, O ) , and E D T A ATPase activity

(A,A) of myosin reacted with a fourfold molar excess of p-phenylenedimaleimide. The dashed line represents the Ca2+ ATPase activity of myosin reacted with a fourfold molar excess of N E M .

S H ] group followed by modification of SH2 group on the SHI-labeled myosin. To exclude such a possibility myosin samples modified with the dimaleimide reagent were diluted after achieving maximum elevation of their CaATPase. The rate of the activity loss (in CaATPase) was found to be independent of the protein concentration in the range of 1-10 mg/ml. Consequently the activity decrease could not be attributed to a separate reaction of the bifunctional reagent with SH2, but rather to an intramolecular process. To eliminate the possibility that disulfide bond formation is responsible for the loss in CaATPase activity of SHI-phenylenedimaleimide myosin, the modified protein was treated (on reaching its elevated CaATPase level) with a 500-fold molar excess of P-mercaptoethanol which was subsequently removed by dialysis. This treatment should block any residual free and monofunctionally substituted dimaleimide precluding any further reaction of this reagent. It should not, however, prevent disulfide bond formation. The above treatment resulted in a n almost complete stabilization of the elevated CaATPase activity indicating that disulfide formation (if it occurs) is not responsible for the loss in CaATPase. It appears, therefore, that the intramolecular reaction which takes place is the bridging by the dimaleimide between the S H ] and SH2 groups of myosin. To substantiate the last conclusion that SH2 groups react with the bifunctional maleimide attached to S H l groups we attempted to show that the loss in CaATPase activity of S H I dimaleimide myosin follows the pattern established for blocking of SH2 groups on S H I - N E M myosin. To achieve this we have isolated S H I-phenylenedimaleimide myosin (prepared at pH 7.0 by reacting the protein for 30 min with an eightfold molar excess of the regent) from the unreacted bifunctional reagent on a G-25 Sephadex column equilibrated at pH 7.0. Following this separation the loss in CaATPase of the purified S H 1-dimaleimide myosin was examined a t pH 7.0, pH 7.9, and in the presence of MgADP (pH 7.9). The results presented in Figure 5 demonstrate that the decrease in CaATPase activity parallels the behavior observed upon blocking of SH2 groups in S H I - N E M myosin (compare also Figure 2). In the present case the modification of SH2 groups cannot be ascribed to excess bifunctional reagent, since the excess unreacted reagent was removed by gel filtration. Thus the loss in CaATPase activity of S H I-dimaleimide myosin arises from the intramolecular bridging of S H I and SH2 groups via the dimaleimide reagent.

In our previous work (Reisler et al., 1974) we have demonstrated that MgADP exposes the SH2 group of native myosin to reaction with N E M while protecting the S H I group from the modification. When the same reaction in the presence of MgADP is performed with the dimaleimide reagent the results obtained are quite revealing. As shown in Figure 4 both Ca and EDTA ATPase activities decrease a t the same rate and to the same extent. Such behavior is consistent with formation of a dimaleimide bridge between S H I and SH2 groups of myosin. The reaction apparently proceeds through blocking of SH2 groups followed by immediate bridging to S H 1. Any conceivable sequential modification of the two groups would lead to separate Ca and EDTA activity curves. Thus it is not surprising that the simultaneous decrease in Ca and EDTA ATPases has not been achieved by modification of myosin with NEM. The latter reagent exhibits preferential specificity toward either S H ] or SH2 group according to the reaction conditions. Our preliminary results demonstrate that the reaction of p phenylenedimaleimide with a proteolytic fragment of myosin, subfragment I, proceeds in identical fashion with that with myosin indicating that the bridges formed between the S H I and SH2 groups reside within each “head” of the myosin molecule and not between two adjacent “heads.” Since the dimaleimide bridge allows for a cross-linking span of 12-14 A, the two essential sulfhydryl groups of myosin are located 12- 14 A apart or, more precisely, may reside temporally within that distance during their interaction with the Mg-nucleotide complex. Moreover, it seems possible that flexibility of some segments of the peptide chain may allow the S H I and SH2 groups to come into an even closer spatial relationship. Experiments employing bifunctional reagents of different cross-linking span should allow the spatial freedom of these groups to be established as well as the range of their closest approach. Such studies including peptide analysis of products of the reaction of myosin with bifunctional maleimide are currently under investigation in our laboratory. The already established proximity of S H I and S H 2 groups is consistent with and supports the previously postulated specific interaction of MgATP with the two sulfhydryl sites to form a ternary cyclic complex. T h e finding that the two sulfhydryl groups S H l a n d SH2 are located with 12-14 A is of considerable interest in view of the recent low resolution structure studies of adenylate kinase (Schulz et al., 1973). It appears from this work that the two thiol groups of the molecule are located a t about this distance on either side of a 10 A deep cleft thought to be the active site region. The two S H groups react at substantially different rates with the fluorogenic reagent 7-chloro-4-nitrobenzo-2oxa- 1,3-diazole (Price, 1972). Reaction of the “fast” group BIOCHEMISTRY. VOL.

1 3 , NO. 1 9 , 1 9 7 4

3839

SCHWAB A N D

with 7-chloro-4-nitrobenzo-2-oxa1,3-diazole leads to complete inactivation of the enzyme, while adenine nucleotide substrates produce a differential effect on the rate of reaction of the two groups reminiscent of the reactivity of the two S H groups of myosin in the presence of A D P and ATP. References Burke, M., Reisler, E., and Harrington, W. F. ( l 9 7 3 ) , Proc. Nat. Acad. Sci. U. S. 70, 3793. Godfrey, J. E . , and Harrington, W. F. ( 1 970), Biochemistry 9, 886. Kielley, W. W., and Bradley, L. B. (1956), J . Biol. Chem. 218, 653. Kielley, W. W., Kalckar, H . M., and Bradley, L. B. (1956). J . Biol. Chem. 219, 95. Mandelkow, E. M., and Mandelkow, E. (1973), FEBS (Fed. Eur. Biochem. Sot.) Lett. 33, 16 I , Price, N.C . (l972), Fed. Proc.. Fed. Amer. S o t . E x p . Biol. 31, 501.

LUKTON

Reisler, E., Burke, M., and Harrington, W. F. (1974), Biochemistry 13, 2014. Schulz, G. E., Biedermann, K. Kabsch, W., and Schirmer, R. H . ( 1 973), J . Mol. Biol. 80, 857. Seidel, J . C. (l969), Biochim. Biophys. Acta 180, 216. Seidel, J. C. ( I 973), Arch. Biochem. Biophys. 157, 588. Sekine, T., Barnett, L. M., and Kielley, W. W. ( 1 962), J . Biol. Chem. 237, 2769. Sekine, T., and Kielley, W. W. (1964), Biochim. Biophys. Acta 81, 336. Sekine, T., and Yamaguchi, M. ( I 963), J . Biochem. ( T o k y o ) 54, 196. Yamaguchi, M., and Sekine, T. (1966), J . Biochem. ( T o k y o ) 59, 24. Yamashita, T., Soma, Y., Kobayashi, S., and Sekine, T. ( 1 965), J . Biochem. (Tokyo) 57,460. Yamashita, T., Soma, Y., Kobayashi, S., Sekine, T., Titani, K., and Narita, K.( 1 964), J . Biochem. ( T o k y o ) 55,576.

Oxidation and Sulfonation of the Highly Reactive Sulfhydryl Groups of Muscle Phosphorylase bt Alfred Schwabf and Aaron Lukton*

ABSTRACT: The reaction of muscle phosphorylase b with a threefold molar excess of o-iodosobenzoate leads to aggregation of the enzyme and a 94% loss of enzyme activity. Substrates and A M P have little or no effect on the inactivation of phosphorylase b by o-iodosobenzoate. Dithiothreitol restores the activity of o-iodosobenzoate-inactivated phosphorylase b to 92% of the control activity and causes reformation of a dimeric structure which is similar to that of the native enzyme. Treatment of oxidized phosphorylase b with 35S022- results in the sulfonation of 1.3- I .6 sulfhydryl groups/molecule of phosphorylase b dimer. The S-sulfonated enzyme has 1.3-2.0 fewer sulfhydryl groups available for rapid reaction with iodoacetamide than native phosphorylase b. Phosphorylase b that has 3.0-3.2 rapidly reacting sulfhydryl groups alkylated, by reaction with iodoacetamide, retains 73-85% of its activity after

reaction with o-iodosobenzoate. It is concluded that those sulfhydryl groups of phosphorylase b that react rapidly with iodoacetamide can be oxidized by o-iodosobenzoate and subsequently sulfonated by reaction with sulfite ion. The inactivation of phosphorylase b by o-iodosobenzoate is attributed to the change in the state of aggregation of the enzyme that occurs when the highly reactive sulfhydryl groups are oxidized. S-SUIfonated phosphorylase b has 59-68% of the control activity and, in the absence of sulfite ion, sediments in the ultracentrifuge as a mixture of dimer (major component) and tetramer. S-Sulfonation of the highly reactive sulfhydryl groups of phosphorylase b leads to a loss in the homotropic cooperativity of A M P sites and a threefold increase in the K , for AMP: these effects can be reversed by dithiothreitol.

B a t t e l l et ai. (l968a) found that the reaction of two sulfhydryl groups of rabbit skeletal muscle phosphorylase b with iodoacetamide, or with various other sulfhydryl reagents, does not cause any loss of enzyme activity. Recently, Zarkadas et a / . (1970) have reported that 3.1 sulfhydryl groups of freshly prepared phosphorylase b can react rapidly with iodoacetamide without loss of enzyme activity. Gold and Blackman ( 1 970) have shown that up to 3.6 sulfhydryl groups of phosphorylase 6 are exceptionally reactive with 2,4-dinitrochlorobenzene and have concluded that these

sulfhydryl groups are identical with those that react rapidly with iodoacetamide. However, dinitrophenylation of the rapidly reacting sulfhydryl groups of phosphorylase b leads to a decrease in the affinity of the enzyme for A M P and glucose 1 phosphate, and a small decrease in the V,,, (Gold, 1968). Apparently, modification of the same sulfhydryl groups of phosphorylase b with different reagents can lead to different effects on enzyme activity. Aggregates of phosphorylase b that disappear upon addition of mercaptoethanol are sometimes seen in the ultracentrifuge (Madsen and Cori, 1956; Seery et al. (1967)). Battell et ul. (1968b) have suggested that such aggregates are due to the formation of intermolecular disulfides by the highly reactive sulfhydryl groups. These investigators also suggested that the reactive sulfhydryl groups may play a role in regulating the ac-

t From the Department of Chemistr), Brooklyn College of the City Lniversity of New York. Brooklyn. New York I1210. Received De( etnher 27, 1973. 1 Present address: Department of Biochemistry, Public Health Research Institute of the City of New York. K e u York. N.Y . 10016.

3840

BIOCHEMISTRY.

VOL

1 3 , ‘4ro 1 9 , 1 9 7 4